Dr BENJAMIN DILDA1, Dr SHAUN CHANDLER1, Dr SHARAD RATANJEE1, Dr EOIN O' SULLIVAN1
1Kidney Health Service, Metro North Hospital and Health Service, Brisbane, Australia, 2QIMR Berghofer Medical Research Institute, Brisbane, Australia, 3Institute for Molecular Bioscience, The University of Queensland, Brisbane, Australia
Biography:
Dr Eoin O’Sullivan is a nephrologist and clinical researcher based in Brisbane, Australia. His work focuses on chronic kidney disease, cellular senescence, and translational bioinformatics, with a particular interest in precision medicine. He leads several investigator-initiated studies and collaborative research programs.
Aim:
To identify appropriate Central Venous Access Device (CVAD) usage rates for incident haemodialysis (HD) patients, identify modifiable drivers of CVAD use, and assess the utility of the Kidney Failure Risk Equation (KFRE2) in guiding locally optimised referrals.
Background:
Although Arteriovenous fistula (AVF) is the preferred access for HD, many patients commence dialysis via CVAD, increasing infection risk and compromising outcomes. Optimising local workflows and understanding drivers of CVAD use is critical for improving vascular access planning.
Methods:
We retrospectively analysed incident HD patients at a large metropolitan centre from 2019–2023, excluding short-term HD. Data were extracted from electronic medical records and ANZDATA. Logistic regression was used to assess predictors of AVF use.
Results:
Of 303 patients, 176 (58%) commenced HD with a CVAD, reducing to 18% at 90 days. Non-modifiable factors (e.g. Acute Kidney Injury, AVF failure, contraindications) accounted for 42% of incident CVAD use – 24% of overall population. Modifiable factors included AVF referral timing (n=25), patient engagement challenges (n=21), and peritoneal dialysis transition (n=56). In our centre, a referral lead-time ≥7.4 months corresponded to 90% AVF use. A KFRE2 >50% was associated with median time to dialysis of 6–7 months, while no patients with KFRE2 ≤20% started dialysis early – suggesting an optimal local protocol. KFRE2 outperformed eGFR in identifying urgent starts.
Conclusions:
A 24% incident CVAD-use limit appears pragmatic based on case-mix analysis. Targeted referral using KFRE2 thresholds may improve AVF rates by enabling earlier intervention. KFRE2 may be especially helpful to avoid false reassurance in moderate eGFRs. Centre-specific lead-time analysis can optimise pathways, and thematic analysis can reveal opportunities for quality improvement initiatives, such as focusing on transitions of care.